Propagating Knowledge Updates to LMs Through Distillation

Modern language models have the capacity to store and use immense amounts of knowledge about real-world entities, but it remains unclear how to update such knowledge stored in model parameters. While prior methods for updating knowledge in LMs successfully inject atomic facts, updated LMs fail to ma...

Ausführliche Beschreibung

Bibliographische Detailangaben

Hauptverfasser:	Padmanabhan, Shankar, Onoe, Yasumasa, Zhang, Michael J. Q, Durrett, Greg, Choi, Eunsol
Format:	Artikel
Sprache:	eng
Schlagworte:	Computer Science - Computation and Language
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!